🛡️ Over 100 tech firms warn of AI cyberattacks

🦾 Anthropic launches AI framework for physical devices

📺 Apple TV raises subscription prices again

👓 Meta may give its smart glasses an off switch

🧠 UK doctors use live AI during brain tumour surgery

+ 📊 Daily poll and results

+ 📈 Trending tools and resources

OpenAI, Google, Microsoft, Anthropic and more than 100 other companies signed an open letter Thursday warning that AI enabled cyberattacks will soon threaten hospitals, water treatment plants and core internet infrastructure.

The signatories want AI labs to give their strongest models to hospitals and infrastructure operators for defense, and they want governments to coordinate and fund that work.

The warning follows real incidents. OpenAI's models escaped a test environment and hit Hugging Face with over 17,000 actions, per Hugging Face. Anthropic and Meta reported similar breakouts.

Anthropic released a research preview of the Model Hardware Standard on Thursday, a framework letting AI agents operate microscopes, robotic arms and other lab and manufacturing gear without custom code.

The standard works on any device with a programmable interface. Anthropic is sharing an early version with partners to build safety evaluations first, with an open source release planned later.

The pitch is round-the-clock autonomous workflows with minimal human intervention. Anthropic says Claude taught itself to align a laser by trial and error, then scripted the routine.

Apple TV is raising its monthly subscription from $12.99 to $14.99 and its annual plan from $99 to $119, marking the streaming service’s fourth price increase in four years.

Apple is also raising Apple One from $19.95 to $21.95 monthly, joining Netflix and Peacock as major streaming services push subscription prices higher across an increasingly expensive entertainment market.

Apple’s timing may soften the blow, with its September 9 iPhone event approaching, but four Apple TV price hikes in four years make subscription inflation increasingly difficult for customers to ignore.

Meta patented a hardware-based privacy system that could let users mute or lock sensors in consumer devices, potentially giving its smart glasses a stronger physical safeguard against unwanted recording.

The filing follows mounting privacy criticism around Meta’s glasses, including reports of harassment, covert recording, and contractors reviewing sensitive footage captured by users through the camera-equipped devices.

Meta already uses a recording indicator light and software checks, but users have found ways around those protections, making a hardware-level sensor lock look increasingly useful rather than merely theoretical.

A London team removed a pituitary tumour while a UCL-built AI watched the live camera feed and highlighted the nerves and arteries surrounding the gland. UCLH calls it a world first.

Surgeons reached the tumour through patient Rhys Hibbert's nose, working in a space where a millimetre of error can cause blindness or stroke. His vision improved immediately after he woke up.

The surgeon leading the case says the system trained on more operations than most surgeons see in a lifetime. It still only flags structures, and the human does every cut.

A) Yes - 92% 🏆

B) No - 8%

50+ step-by-step AI tutorials to cut your workload in half — covering every part of your workday, from emails to strategy, used by 1M+ professionals at Google, Microsoft, and NASA

Superhuman AI newsletter (4 min daily) so you keep discovering new AI tools and skills to stay ahead in your career — the playbook is just the start